Jackery Battery Bank FAIL - Part 2

[@mikeselectricstuff]

Dual back-to-back MOSFETS are standard in this application, so it can shut off both charge and discharge current - need two due to the internal body diode

[@WolfmanDude]

I hate SMD codes so much. Caused so much trouble and wasted time for me. Dammit, just put the full part number on there! Or at least provide a good, free to use databank.

[@EEVblog2]

Yeah, not rocket science to put the number on there.

[@tinygriffy]

With emphasis on "good" .. I tried 3 different smd marking databases with no luck.

[@shawnbergin8479]

the approach I use to all issues of this type is these days is "What is your time worth?", Works for me! Thank you for the work you put in on this.

[@EEVblog2]

Totally.

[@russellhltn1396]

But at the same time, "What is your education worth?"

[@tinygriffy]

you saving an hour .. is it worth your (and my) kids living in a global garbage dump.. I mean it is kind of too late for that in some parts of the world, but still... even if I could earn a 100 Dollars in that 5 hours I need to fix my power bank I think it is totally worth fixing it.

[@JFirn86Q]

It does show up in Mouser. "W-DFN5020-6" is the package name. You were looking at just DFN or something before. It's tricky, but their are a bizzilion package variations so you gotta look at each possible one in the list it could be. The 5020 portion I imagine means 5x2.

[@marcovpv]

Is it normal that they use 5020 for these small parts? I feel like everywhere else 5020 would be 5 cm(50mm) 2 cm(20mm)?

[@pcfreak1992]

@@marcovpv I assume the 0 is a placeholder for decimal digits. So a 5x2.4 would be 5024

[@LemmingGoBoom]

@@marcovpv very common, there's lot of parts that are specced like that.

[@tinygriffy]

Yeah, I found it too... in the heat of the battle ;)

[@ailivac]

And Mouser puts zero effort into normalizing anything (or even making sure all the columns for every part are actually populated). I basically never order from them unless I already know the exact part number ahead of time (and if it's something like a connector I have to double check the entire thing against a manufacturer catalog because of course you can't trust the photos of the few parts that even have them either) because it's so impossible to find anything using their site.

[@bandittwothree3765]

Dave, here's a fix: momentarily short the CS pin to the VSS pin. I've fixed a few packs this way. IIRC, there was an weird voltage on the CS pin, and it's some kind of fault with the chip where it's latched in an in between state, and needs a hi or low to get out of it. I see the Fortune DW01-G version 1.0 2009 datasheet lists this on page 8 in red lettering. I don't know why this works, but it does, and my memory is that I either connected it to a hi or low state, and it reset (through a current limiting resistor for safety, of course.) If you see some odd voltage on the CS pin, I think that's the giveaway.

[@getcartercarpark.]

Very very interesting. I hope you do a part three follow-up with a fix!

[@marks2c]

Dual mosfet is being used as a bi-lateral switch so that batter is protected from both change and discharge.

[@telluridecolorado8918]

Bet if you go to a local "Dollar Store" and find a basic USB power bank with a single 18650, you can probably find that exact same DW01 chip.

[@chitlitlah]

I know from watching Big Clive that the typical equivalent of the Dollar Store in the UK is Poundland. Now you have me wondering what it is in Australia.

[@becejlbe]

Dw01 is on every protected tp4054 board

[@MattyEngland]

ShrimpShop

[@threeMetreJim]

Probably Dave was duped with an expensive brand that only contains standard chinese parts. Most of the cost is the cells, if they are any good. Price is not always an indication of quality.

[@iamdarkyoshi]

I had a powerbank that trashed itself from plugging it into itself. It seemed to use the inductor both as a boost converter for the output and a buck converter for the input. Bigclive did a teardown of one.

[@Youtubecensoredmyusername]

Since you name dropped big Clive. I’m calling it “B sexual” because “a sexual” self reproduce but B sexual destroy themselves upon insertion to themselves 😂

[@akshonclip]

You plugged it into itself? Trying to make perpetual energy device. 😂

[@Youtubecensoredmyusername]

@@akshonclip even if it allowed due to losses you would have to charge back up after a time. That’s hilarious

[@akshonclip]

@@KZbincensoredmyusername Well. It would keep it exercised. 😂

[@luckybob77]

I'm just looking at that thermal sensor. what if it was shorted/open? I'd wager the micro controller would keep everything off out of safety, if it thought it was overheated.

[@EEVblog2]

Thermister is fine.

[@pie314isi]

often the MOSFET RDSon is used as the sense resistor for the overcurrent detect for the cutout. if you parallel them you can choose the current range you want it to work for.

[@LarsBgildThomsen]

Well, you now have "stuff" for a second repair video - your bench power supply ;)

[@EEVblog2]

Don't think I could be bothered unless it gets worse. Still prefectly usable.

[@tinygriffy]

haha.. them repair jobbies keep coming up ;) .. I'd like to see that repair video ! .. guess it is just the LCD connector.. that rubber thingie, but who knows.

[@KeritechElectronics]

Interesting stuff indeed, waiting for part 3 :)

[@RogerTannous]

The 6-pin chip may be similar and compatible to the DW01, but it's not definitely a DW01, because DW01s are simply marked with DW01

[@pursuitforspeed]

Thanks for the follow up 😁 Would not have bet on the controller. Would be the last thing I'd expect. Surprising

[@pglick123]

@11:44 your audio lost sync with the video. With your PC meter fail, it's time to reboot! Everyone is glad you came back to this.

[@Agent24Electronics]

If it has one, you could have a shorted ceramic capacitor on the controller's reset pin holding it in reset or something.

[@EEVblog2]

There is no reset pin.

[@PilotPlater]

Just classic diagnosis. Makes ya feel silly. Good on ya Dave for showing how these things go

[@PaulSteMarie]

Check the thermistor, i think. I also question whether those are simple MOSFETs. I think you have to have back-to-back MOSFETs since the protection circuit has to stop current flowing in when overcharged, and stop current flowing out when overloaded. Some more detailed tracing is in order, i think . BTW, ask BigClive about that DW01 chip. He's reverse engineered dozens of circuits using it. The DW01 should hold *_both_* gates high if no fault is detected. Based on which one is low, you ought to be able to determine what it thinks the fault is, overload or overcharge.

[@nightshadelenar]

the thermistor likely is the case. too cold (too low) or too warm (too high) and it'll lock out.

[@EEVblog2]

Thermistor is fine.

[@EricWAtchesVideos]

I bought a cheap 18650 charger module. It uses a similar control. The chips must be bad as well as I can measure power in, battery, but nothing on the output at all.

[@xani666]

I love how in half of the video MOSFETS were called MooseFETs :D

[@HwAoRrDk]

A MøøseFET once zapped my sister, you know...

[@EEVblog2]

Well known industry term.

[@trevthea5781]

Dave, you actually found the chip family @ 1:51 into your video. It's the DW01A shown in the Google search results. Doh!!!!

[@baconsledge]

Have to say, Dave has the least methodical troubleshooting skills I have ever seen. Watching him analyze his own multimeter almost put me in a coma. Crikey!

[@electrodacus]

The exact mosfets you have are this part number DMG5802LFX and inside there are dual mosfets. Seems you found it :)

[@EEVblog2]

Yep, found it, about 7:30 Looks like they are pretty special snowflake, doesn't seem to be any package equivalents?

[@ronniepirtlejr2606]

Programed to give up the ghost after 5000 cycles. lol

[@EEVblog2]

5000? That's generous.

[@rocketman221projects]

Regular lithium ion cells can't get anywhere near that many cycles.

[@Agent24Electronics]

Wouldn't surprise me. How could you ever prove it without the source code?

[@deepdimdip]

Dave, how do they get away with an unshielded power inductor right over the controller IC? This configuration should ring a bell in mind of whoever designed that board.

[@dine9093]

Pretty sure both gates need to be driven at 3.3v, just because a NFET is backwards doesn't mean it magically becomes a PFET

[@PaulSteMarie]

Yes.

[@TheHuesSciTech]

@James Doe No, not really. Look at the parasitic diodes drawn on most of the diagrams -- if you had only one MOSFET, then even if the MOSFET was off, it would still permit current to pass through the diode. So you need two MOSFETs to be able to block current in both directions. However, I am 99% sure that the controller would turn both on to permit charge/discharge and turn one/both off when in emergency shutdown mode, because you'd never actually *want* the diodes to be carrying the current (0.7V drop on a 3.7V battery is like a 20% power wastage right there). So it's not really correct to say that "one is for charge and the other is for discharge", it's more like "one is to definitively prevent charging and the other is to definitively prevent discharging". A subtle distinction, but one that reveals (as the OP points out here) that it's really weird that only one of the gates was high in Dave's video.

[@volodumurkalunyak4651]

@@TheHuesSciTech actually MOSFET diodes will carry significant current when charging overdischarged battery (could actually happen) or discharging overcharged one (this would probably never happen with such a powerbank)

[@ken830]

That R&S screen issue looks exactly like what happened to my 2008 Samsung 3D TV a couple of years ago. Some issue with the driver, but wasn't worth my time to try to fix it beyond just reseating the flex connectors.

[@EEVblog2]

Yeah, not worth it unless it gets worse. Perfectly usable.

[@ailivac]

Mouser's parametric search is notoriously useless.

[@Agent24Electronics]

It's because the part is catalogued as W-DFN5020-6 and that package is not what he was searching for.

[@ailivac]

@@Agent24Electronics That's my point. Every manufacturer has their own slightly different language (or chooses to use a different equivalent standard e.g. SC-74 vs SOT23-5) and Mouser makes no attempt to unify them anywhere. Or even fill in all the columns on a lot of parts. Or have useful photos. Digikey probably renames the various manufacturer-specific package designations to something like "6-DFN (5.0x2.0mm)", and also has a "Supplier Device Package" column in case you know that "NanoMAX-10S" or whatever is the actual branding for something you're looking for.

[@lui2urco]

Only speculating: If everything is looking good, the problem maybe could be some kind of protection triggering in the protection IC. If there is an undervoltage and the circuit is open to avoid the deep discharge of battery, you cannot charge the battery again because the circuit is open. In this case you need to charge the battery connecting directly the terminals and recover the normal state of the protection IC

[@Flying0Dismount]

The part does come up if you specify the full name of the package "w-dfn5020", but not just "w-dfn"

[@Jedda73]

Maybe a cracked solder joint? Id imagine that battery has had a few falls by now. Could try a bit of pressure on the control chip and see if it springs back to life.

[@michaelfogarty3239]

I dropped my voltage and nearly fell asleep. every thought about doing one audible. still very interesting

[@TimPerfetto]

Thank you

[@EEVblog2]

Thank you for watching.

[@Cameron_D]

Video out of sync from 11:46 still watchable

[@HwAoRrDk]

Not just me then. I was starting to get a bit confused. 😅

[@MattyEngland]

Microcontroller was my thought from the outset tbh. Given that the MOSFETs are in parallel extremely unlikely both would fail. Maybe a drop has caused cracked solder joint on the power to the MC.

[@KillerSpud]

Do a great Scott and make your own boost converter.

[@threeMetreJim]

Big clive will be helpful here. Probably is a dw01 varient. Are you sure the MOSFETs have the same type number and one doesn't have reversed pins for source and drain? Edit: not likely due to the negative being connected as it is - the Chinese always seem to use low side switching. Sometimes just connecting a charger will re-set a protection device if it has gone into protect and 'got stuck'. Disconnecting the batteries will also have a similar effect unless something like a switcher that generates the output voltage has gone short - then it will always go into protect. Isolating the protection side from the switching output/charging part will tell you which part of the circuit has likely failed (track cutting required). Ruled out protection and also overcharge. 4.4V is heading for fireworks! I always disconnect the batteries before even attempting fault finding. The first time you release smoke on something powered by lithium batteries, you'll understand why.

[@junkerzn7312]

Yah, back-to-back mosfets are commonly used as battery-protect devices. You can protect against reverse voltage, you can allow charging but not discharging, or you can allow discharging without charging, or both. And reverse polarity protection as well. Lookup E-FUSEs

[@nightshadelenar]

6:30 these are likely dual mosfet packages, one one way, the other, well yeah. so with two they are beefing up the current. i'd suspect these are 20v 6 or 8 amp jobs driven in a quasi series-parallel arrangment.

[@388sanju]

Hope to see part 3 soon

[@TheLouKou]

Am I the only one with weird audio desync on half of the video? cant seem to locate any other comments...

[@bartgrefte]

Same here, completely out of sync... edit: At the end it looks like about 8 seconds out of sync.

[@ChrisHarringtonMinneapolis]

Yeah, same here.

[@EEVblog2]

Yep, something went wrong. I think it's Davinci, source material is fine.

[@SeanBZA]

Also had a DW01 fail, though helped by the recycled 32700 mystery cells attached to it. Replaced with an other generic BMS board that I had bought in bulk, and also with 10 recycled 18650 cells from old laptop battery packs, so as to get roughly 10Ah capacity in there. Should make the solar lamp work again.

[@RobertBardos]

I have a good mind to send you my ( partially) failed Jackey it would be expensive to ship from USA but would make an interesting troubleshoot. hope you are doing great!

[@Karthor.]

You didn't see it because you only had DFN-6 as package, it's under W-DFN5020-6

[@EEVblog2]

Obviously!

[@erroltheterrible]

That thermistor solder point look SO suspect...

[@nightshadelenar]

one thing to check is hte thermistor, and it's buddy divider resistor. maybe something in there has given up the ghost and could be a really easy and cheap fix.

[@EEVblog2]

Thermister is fine.

[@nightshadelenar]

@@EEVblog2 that is odd, then. Check the reset circut? Past that, i think the controler died due to a surge or something similar. Common if it is plugged into a pc with a not too well protected psu on the 5v side. Need a good board or a fancy pants charge wart with all that jazz to stop it mostly.

[@JarZiz1984]

Did you check the other chip marked "8205A"? I know for a fact that is another dual MOSFET. It looks like your pack has the 2 MOSFETs on the battery terminal and another one in the middle of the board for some reason.

[@artursmihelsons415]

From my experience - sometimes, when reading smd code, zero and letter "O" looks similar.. Same thing with "8" and "B".. That one is too trap, not only for young players.. 😂 In some cases, I need second look with microscope to determine right markings, because quick look with magnifying glass, in some cases, may screw up search.. 😬

[@phinok.m.628]

Well, obviously it'll have to be two mosfets in series, if it's gonna protect against over charging and over discharging...

[@yurikirsanov8763]

You were searching for DFN-6 while this one was a 'DFN-5020'...

[@gnuorder]

Are the MOSFETs in parallel or just have their source connected together?

[@gnuorder]

I should have watched on. You got to it.

[@kjur18]

NTC might be another step. Short it to ground, you can do that when you want no temperature protection. It should work for boost operation of controller, not sure about charge part.

[@Brian-L]

He’s dead Jim.

[@samholdsworth420]

Like a Dodge truck. Dead on delivery guaranteed expired

[@nigozeroichi2501]

I actually was thinking the micro controller from the beginning

[@chitlitlah]

Yeah, me too. I posted it on the last video, but uhh... KZbin must have removed it.

[@ronniepirtlejr2606]

Programed to give up the ghost after 5000 cycles. 😂

[@nigozeroichi2501]

@@ronniepirtlejr2606 with the way many companies are I wouldn't doubt it at all.

[@tinygriffy]

If drain is the bottom pad for both mosfets, your "fuse link" does nothing, because you wired source to source .. or did I miss something there ? 1:35 circumventing the mosfets would be happening if you draw a horizontal line from the B- pin to the top pad of the unpopulated SMD part and solder a wire there.. I'd guess.

[@EEVblog2]

My link shorted out the mosfets.

[@greendragonmakerspace]

Halfway through this and i'm still betting you can bring it to life by shorting bat- to circuit- over the MOSFETS...

[@greendragonmakerspace]

I do a whole 3 hr course on the DW01. It has a whole host of traps for young players. :D

[@EEVblog2]

Watch the end, I short out the mosfets. Still faulty.

[@burra7]

have you tested the button works? Then what is the chip under the ferrite?

[@EEVblog2]

Watch the video, and also part 1, all covered.

[@vbfhru48]

Apart from the value of learning I believe is not worthing the time and effort to fix it. What about buying a 8$ powerbank case and reuse the cells?

[@Starphot]

Jackery is jacked up!

[@dismayer666]

I thought that Jackery is a good brand. Turns out they use all weird and cheapest stuff...

[@malcolmtill]

Is the pushbutton switch ok ?

[@pie314isi]

is the thermistor working? maybe it's cut out because it thinks its over/under temp.

[@michaell8269]

Thermistor is fine.

[@threeMetreJim]

Oh dear, looks like equipment hit by the recent solar storms (power glitches).

[@gshingles]

Did you have an EMP go off in the lab or something? Power bank, power supply, video editing software... better check everything else ;)

[@MattyEngland]

The BBC said it was Trump or Putin to blame.

[@mbak7801]

@@MattyEngland Sounds feasible.

[@michaelmoore7975]

Betcha got one of those in an old power tool battery pack or charging station.

[@1kreature]

Broken NTC wire causing full shutdown?

[@typxxilps]

btw: any solar power updates ?

[@OnurBuyukcaglar]

Actually moosefets are the best type of FETs. 🦌

[@ofoehr]

whats about your coil, its looks like cold soldered

[@willrobbinson]

possible have a count down timer in micro like some dish washer pcb.s deliberate kill pcb after some cycles

[@chasingcapsaicin]

Inductor dry solder through hole VIA connection

[@mhdalessandro]

Didnt follow that bypass bit.. maybe try to short CS and VSS pins of DW01

[@EEVblog2]

I shorted out the MOSFETS, bypassing the entire battery protection circuit.

[@goirikdhar8149]

They are 2 mosfets in one package... and those are in parallel ig.

[@hardscorerockkssss]

dave did u actually measure cells.looks can deceptive sometimes

[@MiniLuv-1984]

Gees the lip sync must be out by 5s or more.

[@EEVblog2]

Oh, yeah, something went wrong there in the 2nd half, I might know why...

[@jhonbus]

Using quotes or clicking "must include" doesn't do anything on Google any more. Google is pretty much useless these days.

[@leodaso2]

Did you test the button?

[@EEVblog2]

Watch the video.

[@electronite7934]

Tc1240?

[@nahimgudfam]

Jackery are pretty junk considering all they do it batteries.

[@clownhands]

Holy flutter echo

[@shawnbergin8479]

Bin chicken fodder!

[@dtec30]

Mosfet cracked

[@willrobbinson]

yep more e wast

[@DirtyRobot]

This is why people pay Northridge Fix good money to fix things. It would have been a three minute job.

[@preston963]

Like I said yesterday move the inductor & probe the controller because it's probably fuhked!... I love how "engeneers" diag arse backwards!

[@EEVblog2]

WTF. Moving the inductor is the hardest option with the least reward. You check the easy stuff first. And blown battery protection IC's are common.

[@mhdalessandro]

Bleep boop peep pee

[@JonnyMac351]

@@EEVblog2 Don't worry Dave he can't even spell Engineer lol

[@j.f.christ8421]

@@JonnyMac351 And "smart" people like you can't figure out what the quotes are for.

[@JonnyMac351]

@@j.f.christ8421 Hahaha Really champ is that so

[@threeMetreJim]

Oh dear, looks like equipment hit by the recent solar storms (power glitches).